Hi All!
Looking at my own photo above and examining the chainring with a 30-power side-illuminated loup, the wear patterns aren't as obvious as they first appeared to me. If you look closely at the photo, you'll see the
side-marking of every other tooth is indeed caused by the inner links...but the
drive wear on every valley and every advancing tooth face correlates with every roller, as one would expect under drive load/torque -- right on the leading face of each engaging-leading chainring tooth.
Hmm.
Sheldon Brown made his case for even numbers on drivetrain sprockets (chainring, cog) here:
http://sheldonbrown.com/chain-life.html....Bu Surly makes their case for odd numbers (
http://surlybikes.com/info_hole/spew/spew_single-speed_gearing_101 ), saying...
...since the wear of the chain on your single speed’s ring and cog isn’t shared by other rings and cogs, as it would be on a multi-geared bike, going with larger rings and cogs spreads the wear over more teeth, so they last longer. Not such a big deal if you’re using our stainless rings, but with aluminum rings, the smaller sizes will need to be replaced more often than the larger ones.[no argument there -- Dan.] To get the longest life out of your ring and cog, use odd-tooth sizes. [Ah! This last part is interesting! -- Dan].
Hmm, again. It sounds like they're advocating large-diameter, odd-odd gearsets (chainrings and cogs). Sheldon would have readily agreed large diameters reduce individual wear, but the odd-odd endorsement would likely put him at uh, odds with Surly.
I'm beginning to think perhaps a 39x17 would make more sense from the standpoint of most even wear distribution (nevermind the recommended/warranteed Rohloff ratios) by virtue of the "hunting tooth" odd-odd engineering principle, just as Netherlands bike maker Idwork endorses (not necessarily the ratio I mentioned above, but the principle).
Reaching back to the days when I rebuilt automotive drivetrains and gearsets, I remember a “Hunting Tooth Gear Set” has tooth counts that are relatively prime (meaning they have no common factors). Any tooth on either gear contacts every tooth on the mating gear before encountering the same tooth again. As a result, wear is spread evenly over all teeth in the gearbox, resulting in maximum overall life. The name comes from any given gear tooth "hunting" through every meshing combination for its starting mate as the gears turn. hitting it once, then hunting again.
Now, the hunting tooth will have a certain frequency, which is...
gear mesh frequency/least common multiple of the numbers of teeth of the two gears
...often this is simply the product of the numbers of teeth, or close to it, which makes life a bit easier. I can remember when Ford got this wrong with some of their pickup trucks in the 1960s/1970s and it resulted in increased wear and a terrible droning at resonant speeds, typically at motorway cruising speeds. The hunting tooth frequency coincided with the vibration spectrum and you'd not only hear it...you could
feel it by way of bone conduction. It was terribly annoying, but also had the odd effect of moving some people to spontaneous tears [<-- as a complete aside to this, various parts of the body do have a resonate frequency, and in the past, the US Army investigated this in experiments that weaponized low-frequency sound and noted a correlation between certain frequencies and loss of bladder or bowel control. There have been a number of studies related to the frequencies of human resonant vibrations as they relate to body parts and extraneous source stimulation; two interesting examples are here:
http://adsabs.harvard.edu/abs/2001SPIE.4317..469B ...and...
http://www.ncbi.nlm.nih.gov/pubmed/9306739 I suspect this may be why some cyclists simply
cannot stand the flux-current-caused vibrations of bicycle dynohubs as they are transmitted through the handlebars from the front hub. See:
http://swhs.home.xs4all.nl/fiets/tests/verlichting/dynamos/theorie/index_en.html for an example].
My experience with hunting-tooth gear sets involved just that -- direct-mesh gearsets. Bicycles are different, and couple their drivetrains with a roller-link chain of various lengths, depending on chainstay length, chainring/cog diameter, and wear (a link is often removed to compensate for chain stretch). Remember, chain links are removed in pairs -- inner and outer -- unless a half-link is employed, making the chain an odd-numbered chain. I'll have to ponder further the effects of even-odd chain link lengths with even-even, odd-odd, even-odd and odd-even tooth counts.
Those interested in a really good discussion of powertrain design would likely enjoy IPT's
Industrial Trades Handbook: Power Transmission Systems, copyright 1988 by IPT Publishing & Training LTD, author Bruce M. Basaraba, ISBN13: 9780920855041, ISBN10: 0920855040. Unfortunately, it seems to be available only in a paper copy and not online.
Fun stuff for those an inquiring nature!
I look forward to seeing how my drivetrain wear progresses, how, and at what rate.
Best,
Dan. (who evidently doesn't have enough to do at 1:30AM and should be in bed, asleep...)
[Edit to correct typo in the ISBN10 number above]
